Tracking Data and Serve Strategy:
Part 1
Peter Tea

Andre Agassi discovered a unique tick about his rival, Boris Becker, which gifted Agassi with seemingly telepathic abilities. Agassi noticed Becker pointed his tongue in the same direction as his intended serve.
It was only in retirement – after this tick had already propelled Agassi to 8 straight victories against Becker – that Agassi finally revealed it. Becker was floored.
Anticipating your opponent's next move is a tactical advantage in any sport. Knowing what to expect in a competition can help settle your nerves and help you attack your opponents' vulnerabilities.
But with tennis played year-round in over 31 countries, on multiple court surfaces and against different opponents, gathering enough insight is a daunting challenge. For example, it's rare to face the same adversary more than a few times within a tennis season.
And the lack of head-to-head encounters makes it difficult to be tactically aware of all your opponent's tendencies. Luckily, ball-tracking data can fill in the gaps of our missing knowledge – and may even discover something as peculiar as the Becker tick.
We begin by knitting together the patterns of player serve tendencies with serve bounce heatmaps. For our study, we looked at data from the 2019 and 2020 Roland Garros tournaments. Let's examine some high-profile players.

What can we garner from the above serve maps on the deuce court? All players have their own imprints – their own serve DNA. Nadal and Federer lean more down T while Djokovic is more balanced in aiming at both corners: Wide and T.
Meanwhile, WTA players aim more towards the body. Interestingly Ash Barty is a partial exception in the deuce court with more wide serves than Swiatek or Halep though less than Djokovic.
The difference in gender profile might be explained by the higher serve vitality in the men's game.
On average the men win more service points – 61% for men versus 55% for women. And aiming corner-to-corner, in addition to velocity differences, may be contributing to this serve advantage.
The average men's player is also taller and has the luxury of serving from a higher position. The elevated vantage point makes it easier to land a serve inside the box. ATP players are either more willing and/or capable of consistently aiming out Wide and down T.
Beyond physical traits, there are surely other reasons why we observe a variety of serve maps. For instance, what about risk tolerance? Every time a server steps up to serve, they must consider the varying levels of faulting risk. To illustrate, let's dive into maps on first and second serves.

You'll notice that most ATP first serves are aimed Wide and T, while the Body direction is more common on second. Aiming corner-to-corner is risky and more likely to land out; by contrast, aiming Body provides more safety room for misdirected serves to still land inside the box.
You may have also noticed that more second serves were aimed down T. With most players being right-handed, aiming for the T in the deuce court would obviously be targeting a returner's backhand side. Gunning for your opponent's backhand is a common strategy.

By splitting the second serve maps against returner handedness, we see clearly that players overtly target the backhand side. This is shown with more wide serves on Deuce and T serves in the ad court against lefty returners, and the mirror opposite for right-handers.
To summarize, players are remarkably predictable on second serve compared with their first in two ways.
First, players feel emboldened and willing to take risks on first serve but feel more risk averse and wary of double-faulting on second. This explains why more second serves land towards Body and away from the two dangerous corners. Second, they like to aim for the returner's backhand and avoid the returner's more deadly forehand weapon